Deepspeech on Android

tachen.cs · May 27, 2019, 3:23am

Does anyone knows how to deploy deepspeech on android? or on a respberry pi?

lissyx · May 27, 2019, 1:38pm

Have you read the documentation ?

tachen.cs · May 27, 2019, 2:09pm

Sorry, I cannot find any reference document to deploy deepspeech on android.

lissyx · May 27, 2019, 2:18pm

github.com

mozilla/DeepSpeech/blob/master/native_client/README.md#android-devices

# Building DeepSpeech Binaries

If you'd like to build the DeepSpeech binaries yourself, you'll need the following pre-requisites downloaded and installed:

* [Mozilla's TensorFlow `r1.13` branch](https://github.com/mozilla/tensorflow/tree/r1.13)
* [General TensorFlow requirements](https://www.tensorflow.org/install/install_sources)
* [libsox](https://sourceforge.net/projects/sox/)

It is required to use our fork of TensorFlow since it includes fixes for common problems encountered when building the native client files.

If you'd like to build the language bindings or the decoder package, you'll also need:

* [SWIG](http://www.swig.org/)
* [node-pre-gyp](https://github.com/mapbox/node-pre-gyp) (for Node.JS bindings only)


## Dependencies

If you follow these instructions, you should compile your own binaries of DeepSpeech (built on TensorFlow using Bazel).

This file has been truncated. show original

tachen.cs · May 28, 2019, 2:43pm

Thank you very much!

lissyx · May 28, 2019, 2:47pm

We also have libdeepspeech on JCenter: https://bintray.com/alissy/org.mozilla.deepspeech/libdeepspeech

And you can rely on higher-level mozillaspeechlibrary from https://github.com/mozilla/androidspeech/ as well

tachen.cs · May 28, 2019, 3:15pm

Hi, lissyx, I also have another question on the import deepspeech.

I know after deepspeech(v0.2), " from deepspeech import Model" is correct, but it’s not suit for deepspeech(v0.1).

Now I am testing deepspeech(v0.1), How can I correctly import Model from deepspeech?

lissyx · May 28, 2019, 3:21pm

Take a look at the examples / bindings in v0.1.1 ?

lissyx · May 28, 2019, 3:22pm

github.com

mozilla/DeepSpeech/blob/v0.1.1/native_client/python/client.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from __future__ import absolute_import, division, print_function
from timeit import default_timer as timer

import argparse
import sys
import scipy.io.wavfile as wav

from deepspeech.model import Model

# These constants control the beam search decoder

# Beam width used in the CTC decoder when building candidate transcriptions
BEAM_WIDTH = 500

# The alpha hyperparameter of the CTC decoder. Language Model weight
LM_WEIGHT = 1.75

# The beta hyperparameter of the CTC decoder. Word insertion weight (penalty)

This file has been truncated. show original

tachen.cs · May 28, 2019, 4:13pm

Sure, Thank you very much.

Rob_Christopher · April 29, 2021, 6:17pm

Hi everyone, is there any update on DeepSpeech on Android, or is there a preferred solution?

ftyers · April 29, 2021, 8:56pm

Some options:

Rob_Christopher · April 29, 2021, 9:03pm

Looks good, thanks